MLRA 111B - Glynwood B-slope Erosion; Northeastern IN
Stephen Roecker and Neil Martin
2017-05-22
1 Introduction
This analysis looks at the field data collected for the following 2017 MLRA project: - MLRA 111B - Glynwood B-slope Erosion; Northeastern IN
Spatially disaggregate the existing SSURGO polygons for Glynwood B-slope map units using ArcSIE, in order to separate different soil erosion phases. The current SSURGO maps join issues at the SSA boundaries, due to different erosion phases. This project is deemed relevant due to the current interest in Soil Health. Distinguishing the difference in erosion phases may have minimal impact on the majority of soil interpretations, but is believed to be significant in distinguishing crop yields.
2 Data Preparation
library(aqp)
library(soilDB)
library(reshape2)
library(ggplot2)
library(gridExtra)
library(knitr)
library(cluster)
library(caret)
library(party)
library(vegan)
library(rgdal)
library(sp)
library(mapview)2.1 Point Data
# projectmapunit data from NASIS
project <- get_projectmapunit_from_NASIS()
project_nodups <- project[!duplicated(project$nationalmusym), c("nationalmusym", "muname")]
# MUPOLYGONs for the Project
gw_pol <- readOGR(dsn = paste0(ownCloud, "glynwood.shp"), layer = "glynwood")
# Soil Survey Areas
ssa <- readOGR(dsn = paste0(geodata, "soils/soilsa_a_nrcs.shp"), layer = "soilsa_a_nrcs")
# Series Extent of Glynwood from SoilWeb
gw_series <- seriesExtent("Glynwood")
# field Data
gw <- read.csv(paste0(ownCloud, "Pts_gnbero_27Jan17.csv"))
gw_sp <- gw
coordinates(gw_sp) <- ~ long + lat
proj4string(gw_sp) <- CRS("+init=epsg:4326")
gw_sp <- spTransform(gw_sp, CRS(proj4string(gw_pol)))
# spatial overlay field data with mupolygons and merge with nasis mapunits
vars <- c("AREASYMBOL", "nationalmu")
gw <- cbind(gw, over(gw_sp, gw_pol)[vars])
gw <- merge(gw, project_nodups, by.x = "nationalmu", by.y = "nationalmusym", all.x = TRUE, sort = FALSE)
# Extract erosion phases from NASIS and combine component and phase
ero_labels <- c("undisturbed", "slight", "moderate", "severe")
gw <- within(gw, {
EroClassFD = factor(EroClassFD, levels = 0:3, labels = ero_labels)
EroClassSIE = factor(EroClassSIE, levels = 0:3, labels = ero_labels)
EroClassFD2 = ifelse(EroClassFD == "severe", "severe", "slight")
EroClassNASIS = NA
EroClassNASIS[grepl("eroded", muname)] = "eroded"
EroClassNASIS[grepl("sev.|severely", muname)] = "sev.eroded"
EroClassNASIS[!grepl("eroded", muname)] = "non.eroded"
soilname2 = soilname
soilname2 = ifelse(soilname2 %in% c("Glynwood", "Morley", "Shinrock", "Rawson", "Mississinewa"), "Glynwood", soilname2)
soilname2 = ifelse(soilname2 %in% c("Blount", "Elliott"), "Blount", soilname2)
soilname2 = ifelse(soilname2 %in% c("Pewamo", "Pandora", "Mermill"), "Pewamo", soilname2)
soilname3 = paste0(soilname2, ifelse(soilname2 == "Glynwood", paste0("-", EroClassFD2), ""))
})
gw <- transform(gw,
rgb = munsell2rgb(mxhue, mxvalue, mxchroma, return_triplets = TRUE)
)
gw_sp <- gw
coordinates(gw_sp) <- ~ long + lat
proj4string(gw_sp) <- CRS("+init=epsg:4326")
gw_sp <- spTransform(gw_sp, CRS(proj4string(gw_pol)))2.2 Spatial Data
The geodata from the Glynwood points was extracted from several rasters at various resolutions. The data using to generate the ArcSIE model came from a DEM with a resolution of 15-feet. The other used came from the 10-meter USGS NED, which was primarily resampled from LiDAR.
# Extract data from rasters
library(raster)
# NW files
fd <- paste0(geodata, "project_data/11FIN/PointDataEval/")
dd <- c("slope10",
"procrv10",
"plncrv10",
"maxcrv10",
"mincrv10",
"relpos_r5",
"wetness_mp"
)
fp <- paste0(fd, "Mosaic_NW_pts/Derivatives/", dd, "/", "w001001.adf")
rs <- stack(fp)
names(rs) <- dd
proj4string(rs) <- CRS("+init=epsg:2965")
gd_nw <- extract(rs, gw_sp, df = TRUE, sp = TRUE)@data
gd_nw <- subset(gd_nw, !is.na(slope10))
# SE files
fp <- paste0(fd, "Mosaic_SE_pts/Derivatives/", dd, "/", "w001001.adf")
rs <- stack(fp)
names(rs) <- dd
proj4string(rs) <- CRS("+init=epsg:2965")
gd_se <- extract(rs, gw_sp, df = TRUE, sp = TRUE)@data
gd_se <- subset(gd_se, !is.na(slope10))
gd15ft <- rbind(gd_nw, gd_se)
rm(gd_nw, gd_se)
write.csv(gd15ft, file = paste0(ownCloud, "geodata_15ft_extract.csv"))
# Region 11 files
subset_rasters <- function(input, output) {
cat(paste0(input, "\n"))
gdal_translate(
src_dataset = input,
dst_dataset = output,
projwin = c(bb[1,1], bb[2,2], bb[1,2], bb[2,1]),
of = "GTiff",
a_nodata = -99999,
overwrite = TRUE,
verbose = TRUE
)
}
warp_rasters <- function(input, output){
cat(paste0(input,"\n"))
gdalwarp(
srcfile = input,
dstfile = output,
te = bb,
s_srs = CRSargs(CRS("+init=epsg:5070")),
t_srs = CRSargs(CRS("+init=epsg:5070")),
r = "bilinear",
tr = c(10, 10),
of = "GTiff",
overwrite = TRUE,
verbose = TRUE
)
}
fd <- paste0(geodata, "project_data/11FIN/sdat/")
dd2 <- c("ned10m_11FIN.sdat",
"ned10m_11FIN_aspect5.sdat",
"ned10m_11FIN_slope5.sdat",
"ned10m_11FIN_cupro5.sdat",
"ned10m_11FIN_cutan5.sdat",
"ned30m_11FIN_mvalleys.sdat",
"ned30m_11FIN_wetness.sdat",
"ned30m_11FIN_z2stream.sdat",
"nlcd30m_11FIN_lulc2011.sdat"
)
dd_names <- c("elev", "aspect5", "slope5", "kp5", "kt5", "mvalley", "wetness2", "z2streams", "lulc")
dd <- paste0(fd, dd2)
input <- dd
output <- paste0(geodata, "project_data/11FIN/glynwood/", gsub(".sdat", ".tif", dd2))
bb <- bbox(gw_sp <- spTransform(gw_sp, CRS("+init=epsg:5070")))
mapply(subset_rasters, input, output)
mapply(subset_rasters, input = paste0(geodata, "project_data/11FIN/nlcd30m_11FIN_lulc2011.tif"), output = paste0(geodata, "project_data/11FIN/glynwood/nlcd30m_11FIN_lulc2011.tif"))
mapply(warp_rasters, input = output, output = gsub(".tif", "2.tif", gsub("30m", "10m", output)))
dd <- output
rs10m <- stack(dd[grepl("10m", dd)])
names(rs10m) <- dd_names[1:5]
rs30m <- stack(dd[grepl("30m", dd)])
names(rs30m) <- dd_names[6:9]
rs10m <- stack(gsub(".tif", "2.tif", gsub("30m", "10m", output)))
names(rs10m) <- dd_names
gw <- as.data.frame(extract(rs10m, gw_sp, df = TRUE, sp = TRUE))
gd30m <- as.data.frame(extract(rs30m, gw_sp, df = TRUE))
gw <- cbind(gd10m, gd30m[, -1])
rm(gd10m, gd30m)
# Save data
save(gw, gw_sp, gw_pol, gw_series, ssa, ero_labels, file = paste0(ownCloud, "Pts_gnbero_27Jan17_geodata.RData"))3 Map Units vs Field Observations
# Load cached dataset
load(paste0(ownCloud, "Pts_gnbero_27Jan17_geodata.RData"))
ssa <- subset(ssa, areasymbol %in% unique(gw$AREASYMBOL))
ssa <- spTransform(ssa, CRS("+init=epsg:4326"))
# Create interactive map
mapView(gw_series) + ssa + gw_spvals2 <- c("EroClassFD", "EroClassNASIS", "nationalmu", "AREASYMBOL")
gw_sub <- gw[vals2]
# Frequency of field observation vs map unit
# Duplicate the data for each REASYMBOL and relabel MLRA
gw_sub2 <- by(gw_sub, gw_sub$nationalmu, function(x) {
x[vals2][1, ]
x[, "AREASYMBOL"] <- "MLRA"
return(x)
})
gw_sub2 <- do.call("rbind", gw_sub2)
gw_sub <- rbind(gw_sub, gw_sub2)
gw_sub$natmuSsaEro <- with(gw_sub,
paste0(nationalmu, "-", AREASYMBOL, "-", EroClassNASIS)
)
test <- xtabs(~ natmuSsaEro + EroClassFD, data = gw_sub)
kable(test, caption = "Frequence by MUSYM-SSA-EROSION")| undisturbed | slight | moderate | severe | |
|---|---|---|---|---|
| 2t6ll-IN009-sev.eroded | 6 | 19 | 18 | 3 |
| 2t6ll-IN053-sev.eroded | 3 | 13 | 20 | 11 |
| 2t6ll-IN075-sev.eroded | 6 | 8 | 4 | 5 |
| 2t6ll-MLRA-sev.eroded | 15 | 40 | 42 | 19 |
| 2t6lm-IN009-sev.eroded | 2 | 10 | 10 | 0 |
| 2t6lm-IN053-sev.eroded | 1 | 6 | 8 | 12 |
| 2t6lm-IN075-sev.eroded | 3 | 1 | 11 | 9 |
| 2t6lm-IN179-sev.eroded | 5 | 12 | 0 | 13 |
| 2t6lm-MLRA-sev.eroded | 11 | 29 | 29 | 34 |
| 2v4bn-IN069-eroded | 4 | 3 | 6 | 4 |
| 2v4bn-MLRA-eroded | 4 | 3 | 6 | 4 |
| 2v4bp-IN179-eroded | 0 | 3 | 0 | 2 |
| 2v4bp-MLRA-eroded | 0 | 3 | 0 | 2 |
| 5jjt-IN035-sev.eroded | 1 | 1 | 2 | 9 |
| 5jjt-MLRA-sev.eroded | 1 | 1 | 2 | 9 |
| NA-NA-non.eroded | 6 | 9 | 11 | 22 |
kable(round(prop.table(test, 1) * 100), caption = "Percent by MUSYM-SSA-EROSION")| undisturbed | slight | moderate | severe | |
|---|---|---|---|---|
| 2t6ll-IN009-sev.eroded | 13 | 41 | 39 | 7 |
| 2t6ll-IN053-sev.eroded | 6 | 28 | 43 | 23 |
| 2t6ll-IN075-sev.eroded | 26 | 35 | 17 | 22 |
| 2t6ll-MLRA-sev.eroded | 13 | 34 | 36 | 16 |
| 2t6lm-IN009-sev.eroded | 9 | 45 | 45 | 0 |
| 2t6lm-IN053-sev.eroded | 4 | 22 | 30 | 44 |
| 2t6lm-IN075-sev.eroded | 12 | 4 | 46 | 38 |
| 2t6lm-IN179-sev.eroded | 17 | 40 | 0 | 43 |
| 2t6lm-MLRA-sev.eroded | 11 | 28 | 28 | 33 |
| 2v4bn-IN069-eroded | 24 | 18 | 35 | 24 |
| 2v4bn-MLRA-eroded | 24 | 18 | 35 | 24 |
| 2v4bp-IN179-eroded | 0 | 60 | 0 | 40 |
| 2v4bp-MLRA-eroded | 0 | 60 | 0 | 40 |
| 5jjt-IN035-sev.eroded | 8 | 8 | 15 | 69 |
| 5jjt-MLRA-sev.eroded | 8 | 8 | 15 | 69 |
| NA-NA-non.eroded | 12 | 19 | 23 | 46 |
Several of counties phased severely eroded, are not dominanted by field observations classified as severely eroded.
4 Accuracy Assessment of the ArcSIE Predictions
cm <- confusionMatrix(data = gw$EroClassSIE, reference = gw$EroClassFD)
print(cm)## Confusion Matrix and Statistics
##
## Reference
## Prediction undisturbed slight moderate severe
## undisturbed 0 0 0 0
## slight 6 8 11 9
## moderate 24 39 53 37
## severe 7 37 25 44
##
## Overall Statistics
##
## Accuracy : 0.35
## 95% CI : (0.2961, 0.4069)
## No Information Rate : 0.3
## P-Value [Acc > NIR] : 0.03521
##
## Kappa : 0.0767
## Mcnemar's Test P-Value : 1.555e-13
##
## Statistics by Class:
##
## Class: undisturbed Class: slight Class: moderate
## Sensitivity 0.0000 0.09524 0.5955
## Specificity 1.0000 0.87963 0.5261
## Pos Pred Value NaN 0.23529 0.3464
## Neg Pred Value 0.8767 0.71429 0.7551
## Prevalence 0.1233 0.28000 0.2967
## Detection Rate 0.0000 0.02667 0.1767
## Detection Prevalence 0.0000 0.11333 0.5100
## Balanced Accuracy 0.5000 0.48743 0.5608
## Class: severe
## Sensitivity 0.4889
## Specificity 0.6714
## Pos Pred Value 0.3894
## Neg Pred Value 0.7540
## Prevalence 0.3000
## Detection Rate 0.1467
## Detection Prevalence 0.3767
## Balanced Accuracy 0.5802test <- as.data.frame(cm$table)
ggplot(test, aes(x = Reference, y = Freq, fill = Prediction)) +
geom_bar(stat = "identity") +
coord_flip()
The accuracy of the current ArcSIE model appears to be low, according to several metrics. None of the classes achieve a positive predictive value of > 50% for the reference data.
4.1 Boxplots
soil_vals <- c("hzthk", "SolumDp", "CaCO3Dp", "claytotest", "firstbtclay", "mxvalue", "mxchroma")
geo_vals1 <- c("SlopeSIE", "ProfCrv", "PlanCrv", "relpos", "wetness")
geo_vals2 <- c("slope5", "kt5", "kp5", "z2streams", "wetness2", "mvalley")
vals <- c(soil_vals, geo_vals1, geo_vals2)
gw_lo1 <- melt(gw, id.vars = "EroClassFD", measure.vars = vals)
gw_lo2 <- melt(gw, id.vars = "EroClassSIE", measure.vars = vals)
names(gw_lo1)[1] <- "EroClass"
gw_lo1$method <- "FD"
names(gw_lo2)[1] <- "EroClass"
gw_lo2$method <- "SIE"
gw_lo <- rbind(gw_lo1, gw_lo2)
ggplot(gw_lo, aes(x = EroClass, y = value)) +
geom_boxplot() +
facet_wrap(~ paste(variable, method), scales="free", ncol = 4) +
coord_flip()
An exploratory analysis shows a considerable amount of overlap exists between the field determined (FD) erosion classes and measurable soil properties. In comparison the FD and SIE (Soil Inference Engine) erosion classes show different patterns within the boxplots, further suggesting that the SIE classes aren’t capturing the field observations accurately.
4.2 Scatterplots
soil_vals2 <- c("hzthk", "SolumDp", "CaCO3Dp", "claytotest", "firstbtclay") # excluded color, only observed a narrow range thus small differences swamp everthing else
vals <- c(soil_vals2, geo_vals2)
test <- gw[, vals]
test_d <- daisy(scale(test), metric = "gower")
test_mds <- metaMDS(test_d, distance = "gower", autotransform = FALSE, trace = FALSE)
test_pts <- cbind(as.data.frame(test_mds$points), EroClassFD = gw$EroClassFD)
g1 <- ggplot(gw, aes(x = hzthk, y = SolumDp, color = EroClassFD)) +
geom_point(cex = 2, alpha = 0.75) +
theme(aspect.ratio = 1)
g2 <- ggplot(test_pts, aes(x = MDS1, y = MDS2, color = EroClassFD)) +
geom_point(cex = 2, alpha = 0.75) +
theme(aspect.ratio = 1)
grid.arrange(g1, g2, ncol = 2)
According to the scatterplot above it appears that only the severe and slight classes are separatable. The moderate erosion class seems to overlap the most with slight. Both the 15-feet and 10-meter DEM derivatives were evaluated, but the results are similar. The overlap in the FD classes is likely due to bias amongst the soil scientists who collected the data.
4.3 Classification Trees
test <- subset(gw, !is.na(EroClassFD))
test_ct <- ctree(EroClassFD ~ ., data = test[, c("EroClassFD", soil_vals)])
plot(test_ct)
confusionMatrix(data = predict(test_ct, type = "response"), reference = test$EroClassFD)## Confusion Matrix and Statistics
##
## Reference
## Prediction undisturbed slight moderate severe
## undisturbed 30 20 11 2
## slight 3 48 5 1
## moderate 3 17 67 27
## severe 1 0 7 60
##
## Overall Statistics
##
## Accuracy : 0.6788
## 95% CI : (0.6229, 0.7311)
## No Information Rate : 0.298
## P-Value [Acc > NIR] : < 2.2e-16
##
## Kappa : 0.567
## Mcnemar's Test P-Value : 1.943e-06
##
## Statistics by Class:
##
## Class: undisturbed Class: slight Class: moderate
## Sensitivity 0.81081 0.5647 0.7444
## Specificity 0.87547 0.9585 0.7783
## Pos Pred Value 0.47619 0.8421 0.5877
## Neg Pred Value 0.97071 0.8490 0.8777
## Prevalence 0.12252 0.2815 0.2980
## Detection Rate 0.09934 0.1589 0.2219
## Detection Prevalence 0.20861 0.1887 0.3775
## Balanced Accuracy 0.84314 0.7616 0.7614
## Class: severe
## Sensitivity 0.6667
## Specificity 0.9623
## Pos Pred Value 0.8824
## Neg Pred Value 0.8718
## Prevalence 0.2980
## Detection Rate 0.1987
## Detection Prevalence 0.2252
## Balanced Accuracy 0.81455 Hierachical Clusters
In order to see if more separation can be achieved amongst the erosion classes a hierachical classifition was peformed. Four unsupervised classes were chosen and manually matched to the FD classes.
test_c <- hclust(test_d, method = "ward")
plot(test_c, labels = gw$upedonid)
rect.hclust(test_c, k = 4)
clusters <- cbind(gw,
test_pts[, 1:2],
clusters = factor(cutree(test_c, k = 4),
levels = c(2, 3, 1, 4),
labels = ero_labels
)
)
xtabs(~ EroClassFD + clusters, data = clusters)## clusters
## EroClassFD undisturbed slight moderate severe
## undisturbed 14 14 6 3
## slight 12 48 22 3
## moderate 7 24 25 34
## severe 4 0 17 695.1 Scatter Plots
g1 <- ggplot(clusters, aes(x = MDS1, y = MDS2, col = EroClassFD)) +
geom_point(cex = 2, alpha = 0.75) +
theme(aspect.ratio = 1)
g2 <- ggplot(clusters, aes(x = MDS1, y = MDS2, col = clusters), main = "test") +
geom_point(cex = 2, alpha = 0.75) +
theme(aspect.ratio = 1)
grid.arrange(g1, g2, ncol = 2)
In comparison the hierarchical clusters have less overlap when viewed along the multidimensional scaled axes, but still does not seem to separate the moderate class.
5.2 Box Plots
gw_lo3 <- melt(clusters, id.vars = "clusters", measure.vars = c(soil_vals, geo_vals2))
names(gw_lo3)[1] <- "EroClass"
gw_lo3$method <- "clusters"
gw_lo1 <- subset(gw_lo1, ! variable %in% c("relpos", "wetness", "SlopeSIE"))
gw_lo <- rbind(gw_lo1, gw_lo3)
ggplot(gw_lo, aes(x = EroClass, y = value)) +
geom_boxplot() +
facet_wrap(~ paste(variable, method), scales="free", ncol = 4) +
coord_flip()
5.3 Classification Tree
test2 <- ctree(clusters ~ ., data = clusters[, c("clusters", soil_vals)])
plot(test2)
confusionMatrix(data = predict(test2, type = "response"), reference = clusters$clusters)## Confusion Matrix and Statistics
##
## Reference
## Prediction undisturbed slight moderate severe
## undisturbed 36 14 1 0
## slight 1 52 10 2
## moderate 0 19 48 23
## severe 0 1 11 84
##
## Overall Statistics
##
## Accuracy : 0.7285
## 95% CI : (0.6746, 0.7778)
## No Information Rate : 0.3609
## P-Value [Acc > NIR] : < 2.2e-16
##
## Kappa : 0.6302
## Mcnemar's Test P-Value : NA
##
## Statistics by Class:
##
## Class: undisturbed Class: slight Class: moderate
## Sensitivity 0.9730 0.6047 0.6857
## Specificity 0.9434 0.9398 0.8190
## Pos Pred Value 0.7059 0.8000 0.5333
## Neg Pred Value 0.9960 0.8565 0.8962
## Prevalence 0.1225 0.2848 0.2318
## Detection Rate 0.1192 0.1722 0.1589
## Detection Prevalence 0.1689 0.2152 0.2980
## Balanced Accuracy 0.9582 0.7722 0.7523
## Class: severe
## Sensitivity 0.7706
## Specificity 0.9378
## Pos Pred Value 0.8750
## Neg Pred Value 0.8786
## Prevalence 0.3609
## Detection Rate 0.2781
## Detection Prevalence 0.3179
## Balanced Accuracy 0.85426 Statistical Modeling
Below several statistical models are evaluated to see if a more accurate model can be developed.
6.1 FD Classes vs DEM Derivatives
test3 <- ctree(EroClassFD ~ ., data = clusters[, c("EroClassFD", geo_vals2)])
plot(test3)
confusionMatrix(data = predict(test3, type = "response"), reference = clusters$EroClassFD)## Confusion Matrix and Statistics
##
## Reference
## Prediction undisturbed slight moderate severe
## undisturbed 0 0 0 0
## slight 0 0 0 0
## moderate 18 12 21 13
## severe 19 73 69 77
##
## Overall Statistics
##
## Accuracy : 0.3245
## 95% CI : (0.272, 0.3805)
## No Information Rate : 0.298
## P-Value [Acc > NIR] : 0.1724
##
## Kappa : 0.0377
## Mcnemar's Test P-Value : NA
##
## Statistics by Class:
##
## Class: undisturbed Class: slight Class: moderate
## Sensitivity 0.0000 0.0000 0.23333
## Specificity 1.0000 1.0000 0.79717
## Pos Pred Value NaN NaN 0.32812
## Neg Pred Value 0.8775 0.7185 0.71008
## Prevalence 0.1225 0.2815 0.29801
## Detection Rate 0.0000 0.0000 0.06954
## Detection Prevalence 0.0000 0.0000 0.21192
## Balanced Accuracy 0.5000 0.5000 0.51525
## Class: severe
## Sensitivity 0.8556
## Specificity 0.2406
## Pos Pred Value 0.3235
## Neg Pred Value 0.7969
## Prevalence 0.2980
## Detection Rate 0.2550
## Detection Prevalence 0.7881
## Balanced Accuracy 0.5481test3 <- cforest(as.factor(EroClassFD) ~ ., data = gw[, c("EroClassFD", geo_vals2)])
varimp(test3)## slope5 kt5 kp5 z2streams wetness2
## 0.032558559 0.014810811 0.006414414 0.004432432 0.016270270
## mvalley
## -0.001873874confusionMatrix(data = predict(test3, type = "response", OOB = TRUE), reference = gw$EroClassFD)## Confusion Matrix and Statistics
##
## Reference
## Prediction undisturbed slight moderate severe
## undisturbed 3 8 4 4
## slight 8 27 24 17
## moderate 15 25 34 15
## severe 11 25 28 54
##
## Overall Statistics
##
## Accuracy : 0.3907
## 95% CI : (0.3354, 0.4483)
## No Information Rate : 0.298
## P-Value [Acc > NIR] : 0.0003594
##
## Kappa : 0.1505
## Mcnemar's Test P-Value : 0.0194217
##
## Statistics by Class:
##
## Class: undisturbed Class: slight Class: moderate
## Sensitivity 0.081081 0.3176 0.3778
## Specificity 0.939623 0.7742 0.7406
## Pos Pred Value 0.157895 0.3553 0.3820
## Neg Pred Value 0.879859 0.7434 0.7371
## Prevalence 0.122517 0.2815 0.2980
## Detection Rate 0.009934 0.0894 0.1126
## Detection Prevalence 0.062914 0.2517 0.2947
## Balanced Accuracy 0.510352 0.5459 0.5592
## Class: severe
## Sensitivity 0.6000
## Specificity 0.6981
## Pos Pred Value 0.4576
## Neg Pred Value 0.8043
## Prevalence 0.2980
## Detection Rate 0.1788
## Detection Prevalence 0.3907
## Balanced Accuracy 0.64916.2 Clusters vs DEM Derivatives
test4 <- ctree(clusters ~ ., data = clusters[, c("clusters", geo_vals2)])
plot(test4)
confusionMatrix(data = predict(test4, type = "response"), reference = clusters$clusters)## Confusion Matrix and Statistics
##
## Reference
## Prediction undisturbed slight moderate severe
## undisturbed 12 3 1 2
## slight 0 31 1 14
## moderate 13 5 63 12
## severe 12 47 5 81
##
## Overall Statistics
##
## Accuracy : 0.6192
## 95% CI : (0.5618, 0.6742)
## No Information Rate : 0.3609
## P-Value [Acc > NIR] : < 2.2e-16
##
## Kappa : 0.4596
## Mcnemar's Test P-Value : 7.988e-08
##
## Statistics by Class:
##
## Class: undisturbed Class: slight Class: moderate
## Sensitivity 0.32432 0.3605 0.9000
## Specificity 0.97736 0.9306 0.8707
## Pos Pred Value 0.66667 0.6739 0.6774
## Neg Pred Value 0.91197 0.7852 0.9665
## Prevalence 0.12252 0.2848 0.2318
## Detection Rate 0.03974 0.1026 0.2086
## Detection Prevalence 0.05960 0.1523 0.3079
## Balanced Accuracy 0.65084 0.6455 0.8853
## Class: severe
## Sensitivity 0.7431
## Specificity 0.6684
## Pos Pred Value 0.5586
## Neg Pred Value 0.8217
## Prevalence 0.3609
## Detection Rate 0.2682
## Detection Prevalence 0.4801
## Balanced Accuracy 0.7058test4 <- cforest(clusters ~ ., data = clusters[, c("clusters", geo_vals2)])
varimp(test4)## slope5 kt5 kp5 z2streams wetness2 mvalley
## 0.016828829 0.022774775 0.001027027 0.001135135 0.009315315 0.196288288confusionMatrix(data = predict(test4, type = "response", OOB=TRUE), reference = clusters$clusters)## Confusion Matrix and Statistics
##
## Reference
## Prediction undisturbed slight moderate severe
## undisturbed 9 1 3 4
## slight 4 40 1 25
## moderate 13 4 61 11
## severe 11 41 5 69
##
## Overall Statistics
##
## Accuracy : 0.5927
## 95% CI : (0.535, 0.6486)
## No Information Rate : 0.3609
## P-Value [Acc > NIR] : 2.413e-16
##
## Kappa : 0.4249
## Mcnemar's Test P-Value : 0.003769
##
## Statistics by Class:
##
## Class: undisturbed Class: slight Class: moderate
## Sensitivity 0.24324 0.4651 0.8714
## Specificity 0.96981 0.8611 0.8793
## Pos Pred Value 0.52941 0.5714 0.6854
## Neg Pred Value 0.90175 0.8017 0.9577
## Prevalence 0.12252 0.2848 0.2318
## Detection Rate 0.02980 0.1325 0.2020
## Detection Prevalence 0.05629 0.2318 0.2947
## Balanced Accuracy 0.60653 0.6631 0.8754
## Class: severe
## Sensitivity 0.6330
## Specificity 0.7047
## Pos Pred Value 0.5476
## Neg Pred Value 0.7727
## Prevalence 0.3609
## Detection Rate 0.2285
## Detection Prevalence 0.4172
## Balanced Accuracy 0.66886.3 Soil series and phases
Try and model the soil components and phases separately. The data is highly unbalanced. Create separate logistic regression models for similar minor components.
#gw$weights <- ifelse(gw$soilname == "Glynwood", 0.1, 1)
gw$gw_severe <- ifelse(gw$soilname3 == "Glynwood-severe", TRUE, FALSE)
test4 <- cforest(as.factor(gw_severe) ~ elev + slope5 + kt5 + kp5 + wetness2 + mvalley + z2streams, data = gw)
sort(varimp(test4), decreasing = TRUE)## slope5 elev wetness2 kt5 z2streams
## 0.0274774775 0.0251171171 0.0157297297 0.0148468468 0.0022162162
## kp5 mvalley
## 0.0009009009 -0.0003063063confusionMatrix(data = predict(test4, type = "response", OOB = TRUE), reference = gw$gw_severe, positive = "TRUE")## Confusion Matrix and Statistics
##
## Reference
## Prediction FALSE TRUE
## FALSE 197 65
## TRUE 15 25
##
## Accuracy : 0.7351
## 95% CI : (0.6815, 0.784)
## No Information Rate : 0.702
## P-Value [Acc > NIR] : 0.1152
##
## Kappa : 0.2464
## Mcnemar's Test P-Value : 4.293e-08
##
## Sensitivity : 0.27778
## Specificity : 0.92925
## Pos Pred Value : 0.62500
## Neg Pred Value : 0.75191
## Prevalence : 0.29801
## Detection Rate : 0.08278
## Detection Prevalence : 0.13245
## Balanced Accuracy : 0.60351
##
## 'Positive' Class : TRUE
## test3 <- glm(as.factor(gw_severe) ~ elev + slope5 + kt5, data = gw, family = "binomial", na.action = na.exclude)
confusionMatrix(data = predict(test3, type = "response") > 0.4, reference = gw$gw_severe, positive = "TRUE")## Confusion Matrix and Statistics
##
## Reference
## Prediction FALSE TRUE
## FALSE 174 49
## TRUE 35 40
##
## Accuracy : 0.7181
## 95% CI : (0.6634, 0.7685)
## No Information Rate : 0.7013
## P-Value [Acc > NIR] : 0.2865
##
## Kappa : 0.2953
## Mcnemar's Test P-Value : 0.1561
##
## Sensitivity : 0.4494
## Specificity : 0.8325
## Pos Pred Value : 0.5333
## Neg Pred Value : 0.7803
## Prevalence : 0.2987
## Detection Rate : 0.1342
## Detection Prevalence : 0.2517
## Balanced Accuracy : 0.6410
##
## 'Positive' Class : TRUE
## summary(test3)##
## Call:
## glm(formula = as.factor(gw_severe) ~ elev + slope5 + kt5, family = "binomial",
## data = gw, na.action = na.exclude)
##
## Deviance Residuals:
## Min 1Q Median 3Q Max
## -1.5512 -0.8366 -0.6028 1.0314 2.4255
##
## Coefficients:
## Estimate Std. Error z value Pr(>|z|)
## (Intercept) -8.953655 2.044889 -4.379 1.19e-05 ***
## elev 0.024425 0.007071 3.455 0.000551 ***
## slope5 0.388811 0.109621 3.547 0.000390 ***
## kt5 0.089968 0.027948 3.219 0.001286 **
## ---
## Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
##
## (Dispersion parameter for binomial family taken to be 1)
##
## Null deviance: 363.39 on 297 degrees of freedom
## Residual deviance: 327.73 on 294 degrees of freedom
## (4 observations deleted due to missingness)
## AIC: 335.73
##
## Number of Fisher Scoring iterations: 4gw$predicted <- predict(test3, type = "response") > 0.4
gw_lo1 <- melt(gw, id.vars = "gw_severe", measure.vars = vals)
gw_lo2 <- melt(gw, id.vars = "predicted", measure.vars = vals)
gw_lo2 <- na.exclude(gw_lo2)
names(gw_lo1)[1] <- "EroClass"
gw_lo1$method <- "FD"
names(gw_lo2)[1] <- "EroClass"
gw_lo2$method <- "GLM"
gw_lo <- rbind(gw_lo1, gw_lo2)
ggplot(gw_lo, aes(x = EroClass, y = value)) +
geom_boxplot() +
facet_wrap(~ paste(variable, method), scales="free", ncol = 4) +
coord_flip()
predfun <- function(model, data) {
v <- predict(model, data, type = "response")
cbind(
p = as.vector(v$fit)
)
}
r <- predict(rs10m, test3, fun = predfun, index = 1:2, progress = "text")
writeRaster(r[[1]], "C:/workspace/severe_erosion.tif", overwrite = TRUE, progress = "text")
r <-predict(rs10m, test4, type='response', progress='text')
writeRaster(r[[1]], "C:/workspace/severe_erosion_cf.tif", overwrite = TRUE, progress = "text")7 Summary
7.1 Issues
- The field data does not confirm the map units phased severely eroded.
- The previous analysis did not exclude the minor components.
- An exploratory analysis illustrated a considerable amount of overlap exists between the field determined (FD) erosion classes and measurable soil properties. However, the FD erosion classes were similarly predictive as the results of a cluster analysis. The classes derived by cluster analysis appeared to overlap slightly and moderately eroded, and were best separated by depth to CaCO3, while the FD erosion classes were best split on A horizon thickness.
- The accuracy of the current ArcSIE model appears to be low, according to several metrics.
- A preliminary analysis found a random forest model to be 20% more accurate than the ArcSIE model.
- Re-delineating the SSURGO map units will most likely result in numerous small delineations.
7.2 Recommendations
- If disaggregating the SSURGO polygons is deemed impractical or inaccurate, re-label the Glynwood B-slope map units (e.g. severely eroded) to the appropriate map unit concept (e.g moderately eroded).
- Evaluate ways to improve the accuracy of the ArcSIE model, or test alternative models such as random forests.
- Conduct a field review May 15th (?) to evaluate the erosion phase concepts and spatial model.
- Evaluate the effect of lumping erosion classes (e.g. slightly and moderately).
- Access yield data from a producer in order to validate the hypothesis that the erosion phases help explain yield variability.